SITUATIONS RESPONSES TO SUICIDE CRISIS - PARTNERS IN PREVENTION: Queensland Centre For Mental Health Research
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONS OPTIMAL CARE PATHWAYS
ACKNOWLEDGEMENTS
We would like to acknowledge the Traditional Custodians of the land on which our
services are located. We pay our respects to the Elders both past and present and
acknowledge Aboriginal and Torres Strait Islander peoples across the State. We
continue to recognise that to Close the Gap we need to work together with Aboriginal
and Torres Strait Islander people, communities, staff and stakeholders to ensure that
we are meeting the needs of the community.
We acknowledge those who experience suicidality and those lost to suicide, and their
families, friends, loved ones, and others who are affected by suicide.
The authors acknowledge the consultation
Disclaimers provided by the QAMH is not a representation of
the full membership base. Responsibility for any
Queensland Government errors of omissions or commissions remain with
The content presented in this publication is the author(s). The QAMH expressly disclaims any
distributed by the Queensland Government liability for damage resulting from the use of the
as an information source only. The State material contained in this publication.
of Queensland makes no statements,
representations or warranties about the Copyright & ISBN
accuracy, completeness or reliability of any
information contained in this publication. The
State of Queensland disclaims all responsibility
and all liability (including without limitation
for liability in negligence for all expenses,
This document is licensed under a Creative
losses, damages and costs you might incur)
as a result of the information being inaccurate Commons Attribution 3.0 Australia
or incomplete in any way, and for any reason licence. To view a copy of this licence, visit
reliance was placed on such information. creativecommons.org/licenses/by/3.0/au
Queensland Police Service © State of Queensland (Queensland Health)
2020
The authors wish to acknowledge the support
and assistance from the Queensland Police You are free to copy, communicate and adapt
Service in undertaking this research. The views the work, as long as you attribute the State of
expressed in this publication are not necessarily Queensland (Queensland Health).
those of the Queensland Police Service and
any errors of omission or commission are the ISBN: 978-0-6487789-4-3
responsibility of the authors.
Suggested citation
Roses in the Ocean Queensland Forensic Mental Health Service,
The views expressed by people with a lived Metro North Hospital and Health Service, and
experience of suicide engaged in consultation Queensland Centre for Mental Health Research.
throughout this project and publication are their 2020. Partners in Prevention: Understanding
own specific perspectives and do not endeavour and Enhancing First Responses to Suicide Crisis
to represent all lived experience perspectives. Situations – Optimal Care Pathways for People
We acknowledge that all lived experience in Suicidal Crisis who Interact with Police or
insights are valuable and important. Paramedics. Brisbane: Queensland Health.
Queensland Alliance for Mental Health This literature review was prepared by:
(QAMH) Dr Katelyn Kerr1,2,3, Clinical Psychologist in
Queensland Alliance for Mental Health (QAMH) is consultation with: Dr Carla Meurk4,5, Dr Elissa
the peak body for the community mental health Waterson4,5, Associate Professor Ed Heffernan4,5
sector in Queensland.
1
Life Promotion Clinic, Australian Institute for Suicide Research and Prevention
2
Day Programs, Toowong Private Hospital
3
Savoir Rooms Specialist Medical Practice
4
Queensland Centre for Mental Health Research
5
Queensland Forensic Mental Health Service, Metro North Hospital and Health Service
PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONS
OPTIMAL CARE PATHWAYS
OVERVIEW
Individuals who experience a suicide crisis often come into contact with police or paramedics. Those who have experienced a
suicide crisis report deficiencies with the existing system, and police and paramedics report that responding to these events
is one of the most challenging aspects of their role. However, little is known about the nature, extent, precipitating factors,
pathways and outcomes of a suicide related call-out, and what responses will most effectively and compassionately meet the
needs of those in crisis. Partners in Prevention: Understanding and Enhancing First Responses to Suicide Crisis Situations,
funded by the Queensland Health Suicide Prevention Health Taskforce, was established in 2017 to address these knowledge
gaps and inform systems enhancements.
Project overview
The Partners in Prevention project encompassed five major initiatives:
DATA LINKAGE
A linked data study about individuals who came into contact with Queensland Police Service or
Queensland Ambulance Service between 2014 and 2017, and their health services use and outcomes
between 2013 and 2018.
SERVICE MAPPING
An integrated service mapping of collaborative services involving police, ambulance and mental health
services up to January, 2018.
PERSPECTIVES FROM LIVED EXPERIENCE
A workshop to gather lived experience perspectives on optimal first responses to suicide crisis situations,
and situations involving a recent bereavement due to suicide.
LITERATURE REVIEWS
Reviews of literature on: optimal care pathways following a suicide-related call to emergency services;
evaluation frameworks for collaborative suicide crisis interventions; and data linkage studies in
suicidology.
KNOWLEDGE, SKILLS, ATTITUDES AND CONFIDENCE OF POLICE
A mixed methods study of knowledge, skills, attitudes and confidence of police in responding to suicide
crisis situations.
Our partners
QUEENSLAND CENTRE FOR
ROSES IN THE OCEAN
MENTAL HEALTH RESEARCH
QUEENSLAND AMBULANCE SERVICE BRISBANE NORTH PHN
QUEENSLAND HEALTH QUEENSLAND ALLIANCE FOR MENTAL HEALTH
QUEENSLAND POLICE SERVICE QUEENSLAND MENTAL HEALTH COMMISSION
1
CONTENTS
OVERVIEW 1
SUMMARY 3
INTRODUCTION 4
METHODS 4
RESULTS 5
Co-responder models 12
Brief contact interventions 15
Short stay safe havens 16
Blended models 19
Culturally appropriate crisis responses 20
Aftercare services 22
DISCUSSION 26
CONCLUSION 26
REFERENCES 27
2 PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONS
OPTIMAL CARE PATHWAYS
SUMMARY
Why we did this
Tailored, effective interventions, that can be delivered during or immediately following the first response to someone in a
suicidal crisis, are vital in order to support a person out of a crisis and prevent future suicide attempts. The predominant health
pathway used by first responders in Australia is transportation to a hospital emergency department. Recent reports have
highlighted significant barriers to care that await those presenting to emergency departments with mental health problems,
including suicidal crises. Individuals with lived experience of suicide are increasingly vocalising the view that emergency
departments are often inappropriate therapeutic environments.
What we did
We undertook a literature review with the following two aims:
1. To examine the existing evidence-base on optimal models of care relevant following a suicidal crisis, for those who interact
with police or paramedics, including models that are relevant for specific populations (e.g., those identifying as Aboriginal
or Torres Strait Islander, veterans); and
2. To synthesize literature and highlight approaches to high quality care that could be established in Queensland.
A comprehensive search of peer reviewed literature, as well as grey literature and articles provided by experts, was undertaken.
What we found
The following six types of models of care were identified:
1. Co-responder models
2. Brief contact interventions
3. Short stay safe havens
4. Blended models
5. Culturally appropriate crisis responses
6. Aftercare services
The review found a number of different models that have been trialled nationally and internationally. However, varied intake
criteria, staffing composition, and evaluation methods make it difficult to compare and contrast services. Many articles did not
elaborate on how services were delivered in a way that would allow replication. Evaluations often did not provide or investigate
outcomes for consumers both short and longer term nor have a comprehensive system of evaluation in place.
Conclusions
Several models of care that may meet the needs of individuals who experience suicidal crisis were identified. While these
services show great promise, there are gaps in the evidence-base relating to these services. Implementation of these services
should be supported by the establishment of comprehensive evaluation frameworks. Further gaps that were identified include
limited tailoring to vulnerable sub-groups and those under the age of 18.
3
INTRODUCTION
Recent figures highlight the tragedy of suicide in Australia, more for admission following assessment and treatment)
with suicide deaths in 2017 sitting at a ten-year maximum (Australasian College for Emergency Medicine, 2018).
of 12.6 deaths per 100,000 persons (Australian Bureau of
Statistics, 2018). Suicide is often preceded by a suicidal crisis The World Health Organization advises that first responders
or crises. attending to someone in suicidal crisis should ensure that
significant others are present and accept responsibility for
Approximately three out of every 100 Australian adults will helping a person in suicidal crisis to gain help, rather than
attempt suicide during their lifetime and more than four out leaving the person alone (World Health Organization, 2009).
of every 1000 Australian adults will make an attempt in any Yet, there is no guidance for first responders on minimal or
one year (Johnston, Pirkis, & Burgess, 2009). non-interventionist approaches that would empower and
allow individuals in crisis to remain safely in situ. While the
First responders play a crucial role in helping people through harms of over medicalising, institutionalising or depriving the
suicidal crises and facilitating appropriate intervention and liberty (through exercise of legislative powers) of those who
follow up. The World Health Organization acknowledges experience mental health problems or suicidal crisis are well
that: “first responders are in a unique position to determine known, the use of legislative powers to involuntarily transport
the course and outcome of suicidal crises” (World Health individuals in crisis to an emergency department are often
Organization, 2009). Tailored, effective interventions that viewed as the only available solution. At present, there are no
can be delivered during or immediately following the first national guidelines on optimal care pathways to facilitate the
response are vital in order to support a person out of a crisis diversion of people in a suicidal crisis from presenting to an
and prevent future suicide attempts. The first response to emergency department (Wilhelm et al., 2007).
a suicidal crisis represents an opportunity to save a life,
but also for first responders to de-stigmatise talking about
suicide, build rapport with the person in crisis, gain trust, and Aims
foster positive public perceptions of their professions.
This current review had two aims:
In the Pathways to Care report (McPhedran & De Leo, 2013),
it was identified that the predominant pathway used by 1. To examine the existing evidence-base on models of
care following a suicidal crisis, for those who interact
first responders for persons who are in, or at high risk of,
with police or paramedics, including models that are
a suicidal crisis is transportation to a hospital emergency
relevant for specific populations (e.g., those identifying
department (McPhedran & De Leo, 2013). The World Health as Aboriginal or Torres Strait Islander; young Australians;
Organization recommends this approach when a person is veterans; and older Australians); and
considered a high or imminent risk to themselves or others
(World Health Organization, 2009). However, when safety 2. To synthesize literature and highlight approaches to high
can be met in other ways, it is recommended first responders quality care that could be established in Queensland.
ensure fast access to mental health treatment (World Health
Organization, 2009). There is limited, if any, peer-reviewed
evidence that transportation to hospital is an optimal
response. METHODS
Recent reports published by the Australian Government For the purposes of this review, we defined a suicidal crisis
Productivity Commission (2019) and Australasian College as encompassing: suicidal ideation, threatened suicide,
for Emergency Medicine (2018) have highlighted significant threatened intentional self-harm, intentional self-harm,
barriers to care that await those presenting to Australian suicidal behaviours, and suicide attempts. First responders,
emergency departments with mental health problems, here, refer to police or paramedics.
including suicidal crises. At the same time, individuals with
lived experience of suicide are increasingly vocalising the view
that emergency departments are inappropriate therapeutic
environments for many of those who are experiencing
crisis (Meurk & Smith, 2020). People with mental health
problems are less likely than those presenting with other
types of problems to be seen in the emergency department
within clinically recommended waiting times (Productivity
Commission, 2019), and are disproportionately likely to
experience access block (defined as waiting eight hours or
4 PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONS
OPTIMAL CARE PATHWAYS
Search strategy Screening and data extraction
The search comprised a database search of peer reviewed Documents were screened for inclusion/exclusion through
literature and grey literature, supplemented by consultation title and abstract screening (title and executive summary
with known experts and snowballing from the original corpus screening for reports). For all documents included, the
of information. following information was extracted: Reference details;
model description (including name, type and staffing);
Academic databases searched were: PubMed, PsycInfo, and
setting (location); consumer characteristics (age, gender,
The Griffith University Library Repository. A grey literature
and characteristics of presentation); evaluation type (if
search was guided by clinical knowledge, and consultation
conducted); period of evaluation; sample characteristics
with known experts to identify candidate care types that
(number of cases/individuals); outcome measures collected;
could comprise a relevant model of care. For each of the
and key outcomes reported.
models identified, the lead author (Dr Katelyn Kerr) sought to
identify peer reviewed or published evaluations, and made
direct contact with service providers to uncover published or
unpublished materials. Analysis
Literature was categorised into care types. An initial
grouping was undertaken by Dr Katelyn Kerr and Dr Carla
Inclusion criteria Meurk, before being verified by all authors and then
discussed among the Partners in Prevention Steering Group.
Following initial scoping of the literature available, we took
Available information for each model was summarised
an iterative approach to identifying and including literature,
individually.
based on an evolving conceptual framework of potential
care pathways for individuals in suicidal crisis who come in
contact with first responders. An initial conceptual framework
was developed that focussed on identifying literature with
respect to the seven domains of: 1. ‘doing nothing’; 2. co-
RESULTS
responder models; 3. alternatives to emergency department Fourteen peer reviewed and grey literature articles, spanning
presentations; 4. involuntary or voluntary transportation the following six categories of care type, met the criteria for
to emergency departments; 5. safe-haven café’s; 6. inclusion:
transportation to the watch-house; and 7. field based brief
interventions. This framework was refined iteratively, through 1. Co-responder model
consultation between Dr Katelyn Kerr and Dr Carla Meurk, 2. Brief contact interventions
based on available literature. Care types were included in this 3. Short stay safe havens
review if they met the following criteria:
4. Blended models
1. they had been developed explicitly for populations in
5. Culturally appropriate crisis responses
suicidal or mental health crises who come in contact with
first responders; OR 6. Aftercare services
2. they were judged by the authors to share important Table 1 provides a summary of models of care and literature
similar features with models that met criteria 1; OR included in this review, in terms of models of care type, study
3. they were judged by the authors to be a potential pathway characteristics, and population of interest. Table 2 provides a
for those who come in contact with first responders. summary of models of care and literature in relation to clinical
outcomes measured, case characteristics and outcomes.
Due to the limited evidence-base, inclusion was based on
relevance. No exclusion criteria were set in relation to study
type or availability of a formal published evaluation. No
quality appraisal of evidence was undertaken as a basis for
inclusion or exclusion of models or literature presented.
5Table 1 Summary of literature, describing models of care, study characteristics and population of interest.
Model name; description Author & Year Staffing Evaluation Type/ Sample characteristics/ Setting Study period Gender Age
Design inclusion criteria
1. Co-responder – mental health clinician co-responds with police or paramedics. May provide advice, conduct mobile assessments, and/or take over the care of a person in suicidal crisis
Review of co-responder models; various Puntis et al. Various Systematic review Various Various Various Various Various
(2018) of descriptive and
qualitative studies.
Psychiatric Emergency Response Team (PAM); Bouveng, Paramedics and mental Descriptive study N=1,580 requests for service, data Stockholm, 2015-2016, 56% No age restrictions.
mental health professionals attend emergency Bengtsson, & health professionals reported on N=1,036 individuals with Sweden 12 months Female. Age range of cohort seen 5-100
calls with paramedics, involving members of the Carlborg (2017) severe mental illness or in suicidal years.
community with severe mental illness or acute crisis.
suicide risk.
Police Ambulance Clinical Early Response (PACER); The Allen Police and mental Pre-post study with N=783 assistance requests for Victoria, 2009-2011, Not No age restrictions reported.
secondary units engaged by police or paramedic Consulting health professionals control group. service for Australia 16 months given.
first responders via police communications centre, Group (2012) individuals in a suicidal or mental
local Crisis Assessment and Treatment Teams, or health crisis.
the police Officer in Charge.
A-PACER; as above Lee et al. (2015) Police and mental Descriptive study N=296 contacts for service for Victoria, 2011-2012, 6 60% No age restrictions reported.
health professionals (mixed methods). individuals in a suicidal or mental Australia months Male.
health crisis.
West Moreton co-responder; mental health Meehan, Brack, Police and mental Pre-post study N=171 individuals; N=226 occasions West Moreton, 2017, 4 49% No age restrictions. 51% aged
clinicians work alongside police. Mental health Mansfield, & health professionals (interrupted time of service. Direct contact with Queensland, months Female. 20-39 years.
co-responder undertakes rapid triage following Stedman (2019) series) N=137. Information on disposition Australia.
on-site assessment by police, advises on referrals following contact available for N=122.
to ED or other services, and undertakes follow-up. Individuals were those in a suicidal or
mental health crisis.
2. Brief Contact Interventions – time limited, structured interventions focussed on problem solving, crisis planning, and linking to other services.
Distress Brief Intervention (DBI); A time limited (14 O’Neill (2018) Stage one delivered Currently underway. Interim findings reported on 1,322 Four trial 2016-2021, Not Originally developed for adults,
days) problem-solving intervention available in by police, paramedics, referrals received up to September, sites across ~4.5 years given. but undergoing expansion in
addition to services as usual. ED or primary care 2018. Individuals in distress who Scotland stages to include 16 and 17 year
professionals; stage come to the attention of police, olds, and scope feasibility for
2 delivered by mental ambulance, hospital EDs, or primary those aged 15 years and younger.
health professionals care.
and peer support
workers with lived
experience.
3. Short stay safe havens – safe, comfortable spaces for people in suicide or mental health crisis to go, as an alternative to presenting to an ED.
Aldershot Safe Haven Service; A walk-in centre Griffiths & Gale Mental health Mixed design. N=4,275 attendances at the service, Aldershot, 2016-2017, Not Indicates the service is for adults
that provides an alternative to ED presentation for (2017) professionals and peer Interrupted time approximately 670 individuals Hampshire, 12 months given. (18 years and over).
adults in crisis. support workers with series of impacts, needing mental health support out United
lived experience. descriptive study of of hours. Kingdom
client satisfaction.
6 PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONSModel name; description Author & Year Staffing Evaluation Type/ Sample characteristics/ Setting Study period Gender Age
Design inclusion criteria
St Vincent’s Safe Haven Café; provides a safe, Price Mental health Cost-benefit N=41 participants who completed Melbourne, 2018, 5 Not Indicates the service is for
supportive and welcoming place for people to Waterhouse professionals; peer analysis a survey. Data on ED presentations Victoria months given. adults.
gain information on support options available to Coopers (2018) support workers with from May-September 2018 were (compared
them, understand more about reactions to crises, lived experience; compared with data from the same with: 5
and facilitate development of therapeutic skills to volunteers. period one-year prior and six months months
manage crises and improve mental health. Free immediately prior. (N=62 participants period 12
tea, coffee and snacks are provided. individuals made 400 visits to the months
café over this period.) Individuals prior; and
were people seeking mental health previous 6
support, including loneliness, months)
personal difficulties, or seeking
social connection.
Brisbane North Safe Space; Safe Space provides – Non-clinical staff None currently. – – – – –
a warm, welcoming, supportive environment that
people can go to when experiencing psychological
distress. Activities are offered.
The Living Room (TLR); A community-based Heyland, Emery, Mental health Descriptive study N=56 visits, N=16 clients participated Chicago, 2015, 8 Not No age restrictions reported.
alternative to ED available to “guests” in an & Shattell (2013) professional and peer in study; Illinois, weeks given. However, findings limited to
emotional crisis. The approach is recovery Heyland & support workers with N=228 visits, 87 individuals in mental United States (clients those 18 years of age or over.
orientated and the TLR environment is arranged Johnson (2017) lived experience. health crisis. of America followed up
like a home living room to maximise guests’ 30 days post
comfort. separation
with service)
4. Blended models - a holistic, multi-factorial model of service, beginning at the point of call and that combines multiple service elements.
Crisis Now; A comprehensive model of care. Model – Mental health None currently – – – – –
is centred around principles and practices of professionals; peer
recovery focussed, trauma-informed treatments, support workers with
use of peer workers, commitment to safety lived experience; non-
and zero suicide, and collaboration with law clinical staff.
enforcement.
5. Culturally appropriate crisis responses - crisis services that focus on the needs of specific cultural groups.
Aboriginal and Torres Strait Islander Suicide Dudgeon, Non-clinical staff. Study design/ N=46 Aboriginal and/or Torres Four trial 2015-2016, Not No age restrictions reported.
Prevention Evaluation Project (ATSIPEP); Milroy, Luxford, methodology not Strait Islander families affected by sites across ~ 12 months given.
Comprises two components: A Critical Response & Holland (2017) described suicide or critical incident, including Western
Stream (CRS) operating state-wide with the suicide or situation where suicide Australia
role of providing assistance to Indigenous is a high risk, murders, or multiple
families following a suicide; and a Community casualty events, that place family, kin
Development Stream (CDS) operating in four and community of the deceased at
sites to upskill communities in suicide prevention elevated suicide risk.
activities and critical responses.
National Indigenous Crisis Response Service – Non-clinical staff None currently – – – – –
(NICRS); Similar to ATSIPEP (above) identifying as (however, informed
Aboriginal and/or by ATSIPEP
Torres Strait Islander. evaluation)
PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONS
7Model name; description Author & Year Staffing Evaluation Type/ Sample characteristics/ Setting Study period Gender Age
Design inclusion criteria
6. Aftercare services - Services that link people in need to appropriate services to prevent suicidality in the future.
The Way Back; A non-clinical, short term support beyondblue Non-clinical staff Mixed design. N=122 referrals, N=87 individuals Darwin, 2014-2015, 40% No age restrictions reported.
service that provides linkages to support networks (2016) with human services Pre-post study who were recently discharged from Northern 18 months Male. 78% under 45 years of age.
in the first three months after discharge. Clients training. (quantitative hospital for a suicide attempt or Territory
can access the program via referral from ED or a and qualitative). suicidal crisis. N=46 interviews with
psychiatric inpatient ward. Descriptive stakeholders.
study of client
characteristics.
PAUSE; A peer support program recently – Peer support workers Currently underway. – – – – –
implemented by Brook RED in Brisbane. The with lived experience.
intervention delivered is individualised and
tailored to the person based on their needs,
including peer support focussed on recovery,
advocacy, linkage to services, and providing
education to carers and family members of the
person.
Green Card Clinic; provide expedited access to Wilhelm et al. Mental health Mixed design. N=456 individuals who presented to Sydney, 1998-2005, 57% No age restrictions reported.
three structured treatment sessions with trained (2007) professionals. Pre-post study St Vincent’s hospital ED for deliberate Australia 7 years Female. Mean age =31.6 years.
clinicians, following deliberate self-harm or of outcomes, self-harm or suicidal ideation.
suicidal crisis. descriptive
study of client
characteristics and
feedback.
Pieta House; a suicide intervention outpatient Surgenor, Mental health Pre-post study. N=432 individuals in suicide crisis Ireland No date 44% No age restrictions reported. Age
service for people who are actively suicidal, who Freeman, & professionals. who were invited to participate pre- range given Male. range 18-74.
self-harm, or who have made a suicide attempt; O’Connor (2015) therapy.
intensive therapy is delivered over a timeframe of
up to 12 weeks.
Veteran Suicide Prevention Pilot – Non-clinical staff None currently. – – – – –
ED: Emergency Department
8 PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONSTable 2 Summary of models of care and literature in relation to clinical outcomes measured, case characteristics and outcomes.
Model Name; Clinical outcomes measured Key Findings - Case characteristics Key Findings - Outcomes
Author & Year
1. Co-responder – mental health clinician co-responds with police or paramedics. May provide advice, conduct mobile assessments, and/or take over the care of a person in suicidal crisis
Review of co-responder No NA • Lack of consistency in evaluation methods, no • Some evidence for reduction in use of police powers with the ride-along model.
models (Puntis et al., randomised controlled trials. • Lack of evidence regarding cost effectiveness or clinical outcomes.
2018) • Unclear whether co-responder model had an
impact on psychiatric hospitalisations.
Psychiatric Emergency No 97% of requests received had high • Of those assessed by PAM, 17% were assessed as high or very high suicide risk.
Response Team (PAM) or medium priority. Reasons for • PAM was well received by stakeholders.
(Bouveng et al., 2017) attendance were: Severe suicide
For the people PAM was in contact with, outcomes were:
threat (36%);
suspicion of severe psychiatric illness • No action or referral to other services (34%); • Job handed over to paramedics (10%);
(25%). • Admitted to psychiatry ED (25%); • Admitted to psychiatric child care (4%);
• Admitted to substance use ED (18%); • Person died by suicide at arrival (0.3%);
• Other or unknown (10%).
Police Ambulance No Reason for attendance were: Of those assessed by PACER:
Clinical Early Response
(PACER) (The Allen
Concern for individual’s welfare
(48%); section 10 apprehensions
• 64% of those assessed by PACER did not
require transportation.
• Proportionally more cases were transported by ambulance than police in the
PACER site, compared with the comparator site (40% police and 45% ambulance
Consulting Group, (37%).
• 43% of those that did not require versus 61% police and 37% ambulance).
2012) transportation had been attended to for
threats of suicide, self-harm or harm to others.
• Proportionally fewer cases were transported to ED at the PACER site in
comparison to the comparator site. (34% versus 84% for police; 79% versus 86%
• Use of force by police decreased in the PACER for ambulance).
area and increased in the comparator site. • Proportionally more cases were transported to a psychiatric facility at the PACER
• 36% of cases required transportation from
the PACER site, compared to 99% in the
site in comparison to the comparator site (30% versus 15% for police; 19% versus
8% for ambulance).
comparator site.
A-PACER (Lee et al., No The two most common reasons for Outcome of A-PACER referral:
2015) referral were: Threatened suicide
• 32% were transported to ED; • 20% did not require further assistance;
(33%); Welfare concerns (22%).
• 22% were referred to another service; • 11% were directly admitted to a psychiatric unit;
• 9% of contacts resulted in a criminal charge.
West Moreton Co- No The two most common reasons for For the 122 people who had direct contact with the co-responder team:
Responder (Meehan et referral were: Threatening suicide/
• 67% remained at the scene; • There was a statistically significant reduction in mental health related ED
al., 2019) self-harm (60%); Situational crisis
• 29% were transported to ED (51% were presentations for the six-month period after the program was introduced.
(22%). 36% of individuals had had admitted); • 23% were transported to hospital under an Emergency Examination Authority
previous contact with a mental health
service. 7% were currently case
• 4% were taken into police custody. (Public Health Act).
managed by a mental health service
at the time of the crisis.
PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONS
9Model Name; Clinical outcomes measured Key Findings - Case characteristics Key Findings - Outcomes
Author & Year
2. Brief Contact Interventions - time limited, structured interventions focussed on problem solving, crisis planning, and linking to other services.
Distress Brief None reported so far. Provisional outcomes (as of • 78% of the appropriate concluded cases • Data available up to 24 May, 2018 indicate overall reduction in distress among
Intervention (DBI) September, 2018): engaged in further DBI support participants of 50% from intake to the end of DBI level 2.
(O’Neill, 2018) • 24% self-reported being under the
influence of alcohol/substances at
point of referral.
Presenting problems reported,
included:
• Stress/anxiety; • Suicidal
ideation;
• Low mood; • Self-harm.
3. Short stay safe havens - safe, comfortable spaces for people in suicide or mental health crisis to go, as an alternative to presenting to an ED
Aldershot Safe Haven No Reasons for attendance: Of those who had previously attended an ED:
Service (Griffiths &
• 56% crisis prevention; • 53% showed a decrease in ED attendance • 28% showed an increase in their attendance at ED.
• 23% ‘social reason’;
Gale, 2017)
following introduction to Safe Haven;
• 13% crisis;
• 7% other. Additional outcomes reported, were:
• 16% reduction in admission to acute inpatient • Reduced police deployments in the area of the Safe Haven (66% of calls resulted
psychiatric beds in the Safe Haven service in deployment versus 43% of calls resulted in police deployment. Qualitative data
catchment area, 12 months post; indicate that this is linked to the ability to refer to Safe Haven cafés.). Service
• 42% drop in calls marked “mental health users rated Safe Haven favourably.
related” from 2013 to 2016 with the
introduction of Safe Haven;
St Vincent’s Safe Haven
Café (Price Waterhouse
No NA • 37% of the 41 people who completed a sign in • Modelling based on ED data suggested that 12% of presentations to the café would
sheet reported they would have gone to ED if have resulted in an ED presentations, had the café not existed (60 presentations
Coopers, 2018) the Safe Haven was not open. This equates to over six months).
151 ED presentations that were diverted in a six-
month period based on the self-report data.
Brisbane North Safe – – –
Space
The Living Room (TLR)
(Heyland & Johnson,
Subjective Units of Distress
Rating (SUDS) (0 no distress –
SUDS score on arrival was 7.7.
Presenting problems: stress + anxiety
• Guests reported an average SUDS reduction of • 94% reported they did not re-attend at ED within the 30-day follow-up.
approximately 2 points on departure.
2017; Heyland et al., 10 highest distress). Measured or depression (62.5%); anxiety (25%).
2013) on arrival and discharge (not Guests at TLR reported the following as beneficial outcomes of engagement:
linked to evaluation).
• Talking with someone about the problem and • Resources and referrals provided by staff (40%);
using problem solving skills (80%); • Other specific person-centred factors, such as goal setting, finding the staff helpful,
• Learning coping skills (67%); liking the comfort of the environment (40%).
• Reassurance of knowing the Living Room was
there if needed (53%);
10 PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONSModel Name; Clinical outcomes measured Key Findings - Case characteristics Key Findings - Outcomes
Author & Year
4. Blended models - a holistic, multi-factorial model of service, beginning at the point of call and that combines multiple service elements.
Crisis Now – – –
5. Culturally appropriate crisis responses - crisis services that focus on the needs of specific cultural groups.
Aboriginal and Torres No 21 families directly helped. 21 families directly helped.
Strait Islander Suicide
Prevention Evaluation Services completed: • 497 telephone calls with services, 72 tele-meetings with services;
Project (ATSIPEP)
(Dudgeon et al., 2017)
• 354 telephone calls with affected families, 45
tele-meetings with affected families;
• Conducted 92 face-to-face meetings with families and community groups.
National Indigenous – – –
Crisis Response Service
6. Aftercare services - Services that link people in need to appropriate services to prevent suicidality in the future.
The Way Back
(beyondblue, 2016)
WHO-5 – World Health
Organisation, Well-Being Index
14% of participants identified as
Aboriginal and/or Torres Strait
• Mean score on exit was 22.5. • Qualitative data indicate that clients were highly satisfied with the service.
(0=worst imaginable well- Islander. N=13 clients completed a
being – 25=best imaginable WHO-5 pre- and post. Mean score on
well-being). intake was 12.5.
PAUSE – – –
Green Card Clinic Centre for Epidemiological Two most common reasons for ED Re-attendance rates were 57% at second session and 34% at third session.
(Wilhelm et al., 2007) Studies Depression Scale presentations among attendees 21% of attenders were able to be followed up. Of these:
(CES-D). (0-60. Higher score were: overdose (66%); suicidal
ideation (17%).
• 67% • 16% reported that they had self-harmed again.
indicates greater depressive
symptoms. Score of 16+ of these stated they had made positive
51% of attenders had a previous lifestyle changes since attending the clinic;
identifies individuals at risk of reported instance of deliberate
clinical depression. self-harm.
There was a statistically significant reduction in symptoms of depression among N=40 participants who attended all three sessions and
FANTASTIC lifestyle checklist • Mean CES-D on intake (n=282)
was 35.7 (s.d.=12.0).
completed the post-test CES-D (mean 17.9, s.d. = 12.9).
(0-50. Higher score indicates
greater control over one’s
lifestyle)
• 95% scored 16 or more, indicating
possible depression.
• 85% scored 23 or more, indicating
significant depression.
Mean FANTASTIC scores 25.9 (s.d. = 7.3).
Pieta House (Surgenor Single item indicator (“I have Pre-treatment scores were: Post-treatment scores were:
et al., 2015) high self-esteem”) rated on
• Mean self-esteem score was 1.76 • Mean self-esteem score was 2.79 (s.d. 1.08); • Mean negative suicidal ideation was 7.77 (s.d. 4.82);
5-point scale (1=low; 5=high) (s.d. 1.07)
• Mean PHQ-9 was 10.87 (s.d. 7.47); • Mean positive suicidal ideation was 13.76 (s.d. 3.66).
Patient Health Questionnaire
(PHQ-9). 9-item scale. Lower
• Mean PHQ-9 score was 18.58
(s.d. 5.77) Changes in scores pre- and post-treatment were statistically significant (pCo-responder models Inclusion to the service, if:
Co-responder programs assist first responders by providing
• Ongoing suicide attempt or severe plans;
advice, conducting mobile assessments, and taking • Urgent acute psychiatric condition which requires clinical
assessment for which delay could cause damage to
over the care of a person in suicidal crisis; enabling the
the individual, others, property, or which could cause
first responders to return to duties, avoiding extended significant deterioration of the condition;
engagement with the person in crisis, and thus increasing
first responder capacity. • Post-partum psychosis1.
Exclusion to the service, if:
Psychiatric Emergency Response Team (PAM),
Stockholm, Sweden • Panic attacks;
• Reaction to crisis (e.g., death of a family member);
Model description
• The person has already died by suicide;
The Psychiatric Emergency Response Team (Psykiatrisk akut • Isolated substance abuse;
Mobilitet – PAM) operates in Stockholm, Sweden, and attends
emergency calls involving members of the community with
• Medically unstable1.
severe mental illness or acute suicide risk (Bouveng et al., These criteria are checked by the prehospital emergency call
2017). PAM is described as representing a unique approach to centre, who act as gatekeepers as to whether PAM
pre-hospital care, that provides the consumer with specialist are engaged1.
assessment and intervention, is time efficient (average
waiting time 15-20 minutes, and average assessment time 1 Over a 12-month period, PAM received 1,580 requests, of
hour 15 minutes), and reduces work load for first responders which they attended 80% (n = 1,254). PAM dealt with 1,036
(Bouveng et al., 2017). Suicide prevention is the main objective. unique individuals, of whom 43% were male, and 56% were
PAM operates from 3pm to 1am daily. PAM is staffed by two female. Individuals ranged from 5-100 years of age. The 18-29
specialist psychiatric nurses and a paramedic, who liaise with year old age group presented the largest demand, at 27% of
police, paramedics, rescue services and the psychiatrist on those in contact with PAM (Bouveng et al., 2017).
call at the psychiatric emergency department, who can provide
Of the 1,580 requests for assistance, 1,392 of these were able
advice and assistance, as required (Bouveng et al., 2017). Staff
to be classified into the following categories:
are chosen based on their high level of specialisation and
clinical experience in suicide prevention1. PAM staff operate • Severe suicide threat (36%);
a modified vehicle, with space for conducting assessments, • Suspicion of severe psychiatric illness (25%);
delivering psychiatric treatment, and transporting consumers
(Bouveng et al., 2017). The vehicle is equipped with mobile
• Acute crisis (18%);
computer systems to access medical records (prior to or • Severe suicide attempt (6%);
during assessment of the consumer), psychiatric medications, • Suspicion of intoxication/overdose (3%); and
breathalyser, and a defibrillator (Bouveng et al., 2017). Calls
are triaged by the emergency services call centre operator,
• Other (12%; Bouveng et al., 2017).
and the vehicle is dispatched in order of priority, with suicidal Evaluation
crises a high priority (Bouveng et al., 2017).
An evaluation of the PAM service was undertaken for
Consumer characteristics the year 2015-2016. No clinical outcomes were reported.
Outcomes of contact, for the 1,036 individuals that PAM was
In the initial implementation of PAM, dispatch was made in contact with, were as follows:
on the basis of clinical judgement; there were no formal
guidelines used to assess which presentations warranted • No action or referral to other services (34%);
attendance by PAM1. Decision-making was based on clinical • Admitted to psychiatric emergency department (25%);
assessment, which included: questioning around presenting • Admitted to substance use emergency department (18%);
problem; consideration of historical factors impacting on the
• Job handed over to paramedics (10%);
suicidal crisis; mental illness history; frequency, duration
and severity of suicidal thoughts, plans, intent; risk versus • Admitted to psychiatric child care (4%);
protective factors and what supports can be utilised in the • Person died by suicide at arrival (0.3%); and
community; and mental status examination1. • Other or unknown (10%).
More stringent criteria for PAM involvement were implemented
in January 2019, with inclusion and exclusion criteria as
follows:
Olof Bouveng, Lead Author Bouveng et al., 2017. Personal communication via email. 8th November 2018.
1
12 PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONSOPTIMAL CARE PATHWAYS
The evaluation reported that PAM was well-received by all Evaluation
stakeholders, and that it may have assisted in reducing
There was no consistency across studies in regards to
stigma associated with suicidal crises and mental illness
evaluation methods. Numerous studies did not measure
(Bouveng et al., 2017).
outcome effectiveness. Of the studies that measured the
Limitations use of police powers (k = 5), the co-responder model was
associated with a reduction in the use of police powers to
No data on clinical outcomes were reported. No pre-data nor detain an individual, when the co-responder model was ride-
control group data provided. along in person assessment (Puntis et al., 2018). A reduction
in psychiatric hospitalisations due to the implementation
Systematic review of co-responder models of the co-responder model was found in three studies,
however hospitalisations were found to increase in another
Model description
three studies (Puntis et al., 2018). Seven studies found that
Puntis et al., (2018) conducted a systematic review of co- consumers were positive with regards a co-responder model,
responder models. They found differences in design, with in comparison to previous experiences they had had with
some models involving remote telephone contact only, and police, however three studies reported that consumers were
others involving mobile teams attending the site in person dissatisfied with lack of follow up and referral pathways
(Puntis et al., 2018). Of the 26 papers utilised, 19 different (Puntis et al., 2018). Nine studies reported on perceptions
co-responder models were found: of the co-responder service, finding that service providers
viewed the service positively, and that police viewed it
• Twelve were ride-along, whereby the police officer and
neutrally compared to service as usual, with the main
mental health clinician attended in the vehicle together.
limitations being availability of the co-responder and
• Five encompassed both a ride-along model and restricted hours (Puntis et al., 2018). Cost-effectiveness was
communication support, whereby the clinician can advise
investigated in three studies. One study found co-responders
police via telephone or police radio remotely. Four of
to reduce costs by 23%, one study found it reduced policing
the five predominantly utilised communication support
remotely, with only serious incidents activating the ride- costs but increased health provider costs, and one study
along unit. found costs increased by less than 1% (Puntis et al., 2018).
• Two services utilised communication support Limitations
remotely as the only method.
The authors identified three major limitations of the studies
Pathway of referral to the co-responder was ascertained for reviewed:
15 models, and included emergency control rooms (k = 2 1. Service user characteristics were not described
studies), directly from police on location (k = 4), emergency adequately, if at all;
control rooms plus directly from police (k = 8), and a direct
2. Co-responder models were often poorly described and
co-responder phone line (k = 1; Puntis et al., 2018). Operation there was a wide variation in the operationalisation of
times were variable, ranging from afternoons and evenings services; and
to 24 hours per day (communication support models only),
3. Effectiveness of co-responder models has not been
seven days per week (Puntis et al., 2018).
rigorously tested.
Consumer characteristics The authors concluded that despite the increased use and
Males were more likely to be referred to the co-responder interest in co-responder models recently, overall there is a
teams than females (range 47% – 77% of referrals received) lack of evidence regarding the effectiveness of such models
(Puntis et al., 2018). Five studies reported mean ages, which and consumer outcomes.
ranged from 34.7 years of age to 40 years of age. One study
reported that 46% of consumers were between 18-39 years Australian co-responder models
of age, while two studies reported that most consumers In Australia, a number of co-responder models have been
were between 35-44 years of age (28%) and 35-54 years of trialled. This review specifically outlines the evaluations of
age (35%). Three studies reported the most common reason Australian co-responder units (“the units”):
for referral to the co-responder as: Suicidal behaviour (two
1. Police, Ambulance and Clinical Early Response (PACER),
studies); or “Bizarre or disorganised behaviour” (one study) Melbourne, Victoria (hours of operation not given)
(Puntis et al., 2018).
2. Alfred Police Ambulance and Clinical Early Response
(A-PACER), Victoria (2pm to 10pm)
3. West Moreton Co-Responder Program, Ipswich,
Queensland (2pm to 10pm, days of operation not given)
13Model description • Section 10 (37%);
The PACER units acted as secondary units engaged by police • Family violence (7%);
or paramedic first responders via police communications • Assist CATT (2%);
centres, local Crisis Assessment and Treatment Teams (CATT),
• Assist ambulance (2%);
or the police Officer in Charge (OIC) (Lee et al., 2015; The
Allen Consulting Group, 2012). Each unit comprised a police • Other (3%; The Allen Consulting Group, 2012).
officer and mental health clinician/s. Some models employed Most consumers (54%) seen by PACER were a registered
the mental health clinician to be located within the police mental health client, and most were also known to police.
station, and others involved the police co-responder picking
the clinician up from the local hospital where they would be A-PACER demographics
located (Lee et al., 2015; The Allen Consulting Group, 2012).
During the six-month trial from November 2011 to May 2012,
The units’ aims were:
296 contacts were made with A-PACER. Of these, 60% were
• To provide onsite mental health assessment in a timely male, and 40% were female (Lee et al., 2015).
manner. This was facilitated by access to background
information via a shared police and mental health Reasons for referral, included:
database; • Threatened suicide (33%);
• To reduce unnecessary transportation to hospital • Welfare concerns (22%);
by police or ambulance, through assessing the
appropriateness of community referral options or not
• Psychotic episode (18%);
requiring further intervention; and • Assist CATT or police (12%);
• To assess and facilitate the most appropriate referrals • Family violence (7%);
to those encountered, including to psychologists and • Revoked community order (3%); and
psychiatrists, general practitioners, community based
organisations, as well as hospital and psychiatric facilities • Follow up after prior contact with A-PACER (2%; Lee
et al., 2015).
(The Allen Consulting Group, 2012).
West Moreton co-responder program demographics
Consumer characteristics
Of 137 people who had direct contact with the unit, 50% were
Inclusion criteria for attendance by the PACER units were:
male. Just over one-half (51%) of consumers were between
• Onsite clinical assessment of the crisis was required (The 20-39 years of age (Meehan et al., 2019).
Allen Consulting Group, 2012);
• Onsite or telephone advice was needed regarding referral The most common reasons for co-responder involvement with
options (The Allen Consulting Group, 2012); the 137 who had direct contact were:
• Guidance on transport options and de-escalation • Threatening suicide/self-harm (60%);
techniques was requested (The Allen Consulting Group, • Situational crisis (22%);
2012);
• Threatening harm to others (5%); and
• Assistance in treatment planning for frequent users of • Other not reported (13%).
emergency services was requested (The Allen Consulting
Group, 2012); Personality disorder was the most common diagnosis (22%)
with intoxication accounting for 11% of call outs. Previous
• The clinician had triaged the call and ascertained it was
contact with mental health services had occurred in 36% of
appropriate for attendance (Lee et al., 2015).
cases (Meehan et al., 2019).
Demographics of the units are reported below.
Evaluations
PACER demographics
PACER
During a 16 month period (2009-2011), PACER received 783
assistance requests (The Allen Consulting Group, 2012). No PACER was evaluated over a 16-month period (1 December,
demographic data were reported on. 2009–31 March, 2011; The Allen Consulting Group, 2012). The
evaluation design was a pre-post study with control group.
Reasons for referral, included:
• Welfare concerns (48%, including threatening suicide/ • Over three-quarters (78%) of assistance requests
were responded to on-site. Of this 78%, 82%
self-harm, threatening harm to others, at risk of harm,
received a PACER mental health assessment;
frightening/delusional behaviour, confused/incoherent);
• In addition, 36% of cases required transportation from
the PACER site, compared to 99% in the comparator site;
14 PARTNERS IN PREVENTION: UNDERSTANDING AND ENHANCING FIRST RESPONSES TO SUICIDE CRISIS SITUATIONSOPTIMAL CARE PATHWAYS
• Ambulance transport was used more in PACER (45%) and There was a statistically significant reduction in mental
police transport was used less (40%), compared to the health related emergency department presentations for
comparator site, where ambulance transport was used in the six months after the program was introduced (299
37% of transports and police used in 61% of transports; presentations per month pre-program, and 265 per month
• At the PACER site, 52% of transportations were to hospital post-program; Meehan et al., 2019). Surveys of police first
emergency departments, and 27% were to a psychiatric responder intentions as they were leaving the scene indicated
facility. At the comparator site, 82% of transportations that police had intended to transport 82% of consumers to
were to a hospital emergency department and 12% were hospital under an Emergency Examination Authority (EEA).
to a psychiatric facility (The Allen Consulting Group, 2012). However, when co-responders were present, only 23% of
consumers were transported on an EEA (Meehan et al., 2019).
PACER was well accepted by stakeholders, including police,
paramedics, mental health staff, and viewed favourably by Limitations
consumers who experienced reduced waiting times, more
flexibility in options for care, less need for transportation PACER
to hospital, and better utilisation of resources. The
The evaluation by The Allen Consulting Group (2012) did
authors concluded PACER was more cost effective than the
not report on demographics of consumers. The study
comparator site (The Allen Consulting Group, 2012).
authors noted some limitations due to the fact that police
A-PACER and ambulance service staff did not always complete the
Mental Disorder Transfer Forms (L42), the study’s key data
Of the 296 contacts, 32% were transported to emergency source. Additionally, the evaluation only reported descriptive
department; 22% were referred to a suitable service; 20% statistics, thus limiting the conclusions that can be drawn
were judged as not requiring any further assistance; 11% regarding the efficacy of PACER (The Allen Consulting Group,
were directly admitted to a psychiatry ward; and 9% received 2012).
phone assistance (Lee et al., 2015). Transportation was more
often with police (58% of those needing transportation) than A-PACER
ambulance (36%) or other services (6%; Lee et al., 2015).
No clinical outcome data was collected. No pre- and post-
It was noted by the lead author that PACER/A-PACER models tests for consumers nor questionnaire for consumers on how
function most efficiently in high density metropolitan areas they experienced A-PACER were reported.
where locations are easily reached by car2. Rural areas
West Moreton co-responder program
provide time restraint challenges, as one job may take hours
to respond to2. In these areas, Lee recommended a telephone It is unknown whether service users engaged with services
model could be implemented whereby a clinician could to which they were referred, as this information was not
provide advice and assistance via telephone and mental gathered or available (Meehan et al., 2019). Additionally,
health trained police officers could respond in person to the there has been no follow-up with consumers, nor collection
crisis2. Alternatively, the team could service the town centre of clinical outcomes data. No comparator site evaluation was
in larger rural areas, with police being the respondents to out conducted.
of town calls with telephone assistance from the clinician2.
West Moreton co-responder program Brief contact interventions
During the initial 16 weeks of the program, in 2017, the Brief Contact Interventions, such as the Distress Brief
co-responder team had direct contact with 137 individuals, Intervention (DBI), are time limited, structured, interventions
as well as providing phone advice to other agencies for 44 that aim to deliver a compassionate and proportionate first
people (Meehan et al., 2019). Information for analyses was response to individuals in crisis (The Scottish Government,
available for 122 of the persons who had direct contact with 2015). Delivery of these interventions should not be reliant
the co-responder team. Of these, 67% remained at the scene; on mental health or allied health professionals, but open to
29% were transported to the emergency department; and 4% suitably concerned and interested individuals who receive
were taken into police custody (Meehan et al., 2019). appropriate training. These models of contact are time
limited, but involve ongoing contact, employing multiple
Those who remained at the scene were referred to various
forms of communication (e.g., telephone or written contact).
government and non-government agencies for follow up
(Meehan et al., 2019).
2
Stuart J Lee, Lead Author Lee et al., (2015). Personal communication via email 21st November 2018.
15You can also read
